17 May
2007
17 May
'07
9:37 p.m.
On 5/17/07, Roger Merchberger <zmerch-cctalk at 30below.com> wrote:
My main point for that part of the thread is that
making a source-code
translator from different processor families that would be even "mostly
successful" would be at least 1, probably 2 orders of magnitude more
difficult, IMHO, and would probably have never existed, even if it were
possible.
I will agree that the specific case of 8080->68000 would be difficult,
owing to the differences in their fundamental architecture, but I
can't agree to the general case that one can't write a translator from
one family to another. At Software Results, before I started there,
the initial HASPBOX line of products was all written in MACRO-11
because the underlying comms hardware was 100% DEC (either an 11/04 or
11/23, depending on the year it was made). When SRC moved from a
PDP-11 OEM arrangement to a self-designed 3rd-party 68000-based single
board (the COMBOARD-I), my predecessors used TECO to mung the code
from MACRO-11 to 68000 assembler. I am not claiming
that the process
did not require any manual fixup, but all the heavy lifting was
done
by editor macros, not even a dedicated app that might claim to
"understand" both architectures. I would call that effort "mostly
successful" from a commercial standpoint at least. One can argue that
the PDP-11 and the 68000 share some underlying design philosophies,
but they are certainly not in the same processor family.
So who's gonna buy a wholly offtopic yet o-so-kewl
new Amiga when they're
out? ;-)
Um... not I. Too pricey and not Amiga-y enough.
-ethan